Production in a desert lizard as a consequence

Production in a desert lizard as a consequence of prey availability and annual variation in climate R. Anderson 1, L. Mc. Brayer 2, C. Fabry 1, and P. Dugger 1 1 Western Washington University, 2 Georgia Southern University

Introduction 1 Trophic interactions in desert systems are presumed to be strongly linked. It is reasonable to hypothesize that the annual trophic patterns in desert scrub communities are strongly influenced by annual variation in temperature and precipitation. That is, bottom-up effects in production, from plants to herbivores to secondary consumers carnivores and tertiary consumers (mesopredators and apex predators) should be evident, commensurate with short term effects of climate. We tested the foregoing hypothesis by analyses of 1) individual lizard body condition, 2) lizard abundance from year to year, 3) annual productivity of the lizard’s prey, and annual, short-term climatic patterns in temperature and precipitation.

Subject Animals • Apex, mesopredator as “Tertiary” consumer*: – Leopard Lizard, Gambelia wislizenii • Insectivores as “Secondary” consumers*: – Western Whiptail Lizard, Aspidoscelis tigris – Desert Horned Lizard, Phrynosoma platyrhinos • Insects as “primary” consumers*: – Grasshoppers, cicadas, termites, ants, and more *obviously, trophic levels are mixed for many animals 2

3 Male leopard lizard, Gambelia wislizenii in classic ambush predation pose. It eats large arthropods, especially grasshoppers, and other lizards.

4 Grasshoppers on foliage Grasshoppers in the open Prey of the leopard lizard, Gambelia wislizenii Western whiptail lizard Aspidoscelis tigris Desert horned lizard Phrynosoma platyrhinos

5 Marked female Gambelia wislizenii eating western whiptail lizard, Aspidoscelis tigris.

Research Site Ø Alvord Basin, Harney Co, OR Ø BLM administered public land Ø Great Basin desert scrub Ø 20% cover by perennial vegetation ØMix of sandy flats, dunes, and hardpan mesohabitats Ø Dominant perennial shrubs: • Basin big sage, Artemisia tridentata • Greasewood, Sarcobatus vermiculatus 6

7 On plot, view northward of Alvord Basin, with Steens Mountain, June 2011. (note the extensive cheatgrass in foreground)

Methods • Research period ~June 25 to July 16, 2003 -2011 • Standard plot surveys for grasshoppers • Standardized annual pitfall trapping • Annual census of lizards on a 4 ha core plot • Capture-mark-release of more lizards near plot • Weather records in the field, greatly buttressed from weather station in nearby Fields, OR, compiled by the DRI, under auspices of WRCC. 8

Monthly mean daily air temperatures near study site (Fields) and other weather stations 30 25 Mean Temperature (°C) 20 15 Values are means for the last decade 10 Bly 4 SE Hart Mountain 5 Mc. Dermitt Paradise Valley Fields 0 -5 Rome 2 NW Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 9

10 8 7 Month to month precipitation patterns near study site (Fields) and at other weather stations in the region Mean Precipitation (cm) 6 Values are means for the last decade 5 Bly 4 SE Hart Mountain Mc. Dermitt 4 Paradise Valley Fields 3 Rome 2 NW 2 1 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Four dominant grasshoppers on plot, 2003 -2009 11 Proportion (mean) Trimerotropis pallidipennis 0. 51 Cordillacris occipitalis 0. 29 (Pallid winged gh) (Spotted winged gh) Melanoplus rugglesi 0. 09 (Nevada sage gh) Parapomala pallida 0. 07 (Mantled toothpick gh) Sample sizes: Values are means from 3 counts for each for 2 -3 time periods, for each of 9 days on eight 5 mx 5 m quadrats, with 8 quadrats per 10 mx 40 m plot , 3 plots per meso habitat, per two mesohabitats.

12 Annual variation in number of grasshoppers counted on plot in early July, as related to air temperatures in prior months GH observed per site visit 20 2004 2003 15 Pearson's Corr = 0. 909 p (2 -sided) = 0. 012 10 2008 5 20 2004 15 2003 10 2009 5 2006 2007 0 -6 -5 -4 -3 Dec-Mar mean min temp (°C) -2 -1 0 19 20 21 22 23 24 May mean max temperature (°C) Sample sizes: Values are means from 3 counts for each for 2 -3 time periods, for each of 9 days on eight 5 mx 5 m quadrats, with 8 quadrats per 10 mx 40 m plot , 3 plots per meso habitat, per two mesohabitats. 25

13 Annual variation in number of grasshoppers counted on plot 20 GH observed per site visit 2004 15 2003 10 2009 2008 5 2006 Mean May Rainfall 1998 to 2011 2007 0 0 1 2 3 4 May precipitation (cm) Sample sizes: Values are means from 3 counts for each for 2 -3 time periods, for each of 9 days on eight 5 mx 5 m quadrats, with 8 quadrats per 10 mx 40 m plot , 3 plots per meso habitat, per two mesohabitats. 5

14 0, 110 2004 Log 1 + (Male GW body mass / SVL) 2005 2008 0, 100 2006 2003 2009 p = 0. 045 0, 090 Mean Daily Max among years 2007 0, 080 18, 0 20, 0 22, 0 24, 0 26, 0 28, 0 May mean daily maximum air temperature o. C Linear regression of log-transformed body mass to snout-vent length ratio of Gambelia wislizenii for each year during the 2003 -2009 summer field seasons relative to the mean daily maximum temperature during the preceding May. Numbers for G. wislizenii body mass/snout-vent length ratio were transformed by adding 1, then taking the log of each data point [log(1+(GW body mass/SVL)) = 0. 159 – 0. 00289(Temperature)].

Body condition of male Gambelia wislizenii as presumed function of availability of its primary arthropod prey 0, 28 15 2004 2005 Male GW body mass / snout-vent length 2006 2008 0, 26 2003 2009 0, 24 0, 22 2007 Linear regression of male Gambelia wislizenii (GW) body condition (g body mass per mm snout-vent length) as a function of log-transformed grasshopper-and-cicada availability (log of the sum of the mean number of grasshoppers per site visit and 6 times the mean number of cicadas per site visit, assuming 6 grasshoppers per cicada by weight) per site visit, for each year during the 2003 -2009 summer field seasons. (log(GH+Cicada) = 0. 0480 (GW body mass/SVL) + 0. 212). 0, 20 p = 0. 017 0, 18 0, 00 0, 20 0, 40 0, 60 0, 80 Log (GH + Cicada) 1, 00 1, 20 1, 40

16 Spearman Rank Analysis of factors affecting lizard body condition Male Gw Mass/SVL G-hopper Counts G-hopper + May weather 0. 249(5) 13. 9(2) 5(1) 0. 275(1) 18. 7(1) 0. 258(3) May Max Temps May Rain Winter Min Temps 19. 9(1) 2(1) -2. 27(1) 8(2) 20. 1(2) 5(2) -2. 57(2) 5. 1(5) 17(4. 5) 22. 5(4) 9(4. 5) -4. 37(3) 0. 212(6) 1. 8(6) 23(6) 24. 0(6) 12(6) -5. 02(5) 0. 259(2) 5. 4(4) 13(3) 20. 3(3) 5(3) -5. 34(6) 0. 250(4) 5. 9(3) 17(4. 5) 23. 2(5) 9(4. 5) -4. 61(4) rs 0. 901* 0. 887* 0. 868* 0. 813 0. 890* Asterisks denote significant correlations at N = 6 and α = 0. 05 (r s > 0. 829).

17 Patterns of Arthropod Abundance in Pitfall Traps 2004 -2011 Analysis of Variance* Source Type III SS df Mean Squares F-ratio p-value Year 357, 964. 706 7 51, 137. 815 75. 328 0. 0001 Mesohabitat 31, 120. 345 2 15, 560. 172 22. 921 0. 0001 Plant Species 10, 577. 248 15. 581 0. 000 Plant Size 2, 503. 398 2 1, 251. 699 1. 844 0. 159 Error 494, 893. 417 729 678. 866 *Post hoc tests revealed these significant differences in annual abundances: Higher in 2005, 10, and 11 relative to 2004 and 2006 -09 Rainfall total in both May 2010 & 2011 were about 3. 75 cm

18

19

Among year patterns of 1 yr olds recruited to predator and prey populations (see fig 19) #1 yr #Older Gw #1 yr #Older At #1 yr 20 #Older Pp 2004 6 66 24 75 8 42 2005 38 115 10 77 17 31 2006 8 131 8 41 17 31 2007 6 69 7 30 18 35 2008 1 104 1 64 3 43 2009 4 79 5 37 13 45 2010 24 101 8 97 21 27 2011 19 85 7 70 24 31 14% /yr 14%/yr 42%/yr

Conclusions 21 • Short term climatic extremes in both the inactivity season and activity season may have a direct effect on arthropod prey abundance. • Short term climatic variation in temperature and rainfall results in similar temporal variation of productivity at the lower trophic levels. • Productivity at the lower trophic levels affect productivity at higher trophic levels. • Higher temperatures during daily and seasonal activity periods may have debilitating energetic consequences for mesopredators and apex predators in seasonal environments, particularly if precipitation is low and the bottom-up trophic energy flow is slowed. • More detailed and integrative analyses of the population dynamic patterns of the mesopredator, its vertebrate prey, and their prey may provide further insights to desert trophic interactions. • See the last figure (panel 23) for summary of the interactions

Flowchart of hypothesized and observed abiotic and biotic trophic interactions in the Alvord Basin desert scrub. Arrow size denotes relative strength of observed effects 22